Millimeter-Size Pickering Emulsions Stabilized with Janus Microparticles

Article

Cite This: Langmuir 2019, 35, 4693−4701 pubs.acs.org/Langmuir

Millimeter-Size Pickering Emulsions Stabilized with Janus Microparticles † ‡ § ∥ † Bobby Haney, Dong Chen, Li-Heng Cai, David Weitz, and Subramanian Ramakrishnan*, † Department of Chemical and Biomedical Engineering, FAMU-FSU Engineering, Tallahassee, Florida 32310, United States ‡ College of Chemical and Biological Engineering, Zhejiang University, Zhejiang 310027, China § School of Engineering and Applied Science, University of Virginia, Charlottesville, Virginia 22903, United States ∥ Physics, Harvard University, Cambridge, Massachusetts 02138, United States

ABSTRACT: The ability to make stable water-in-oil and oil-in-water millimeter-size Pickering emulsions is demonstrated using Janus particles particles with distinct surface chemistries. The use of a highly cross-linked hydrophobic polymer network and the excellent water-wetting nature of a hydrogel as the hydrophobic and hydrophilic sides, respectively, permit distinct wettability on the Janus particle. Glass capillary microﬂuidics allows the synthesis of Janus particles with controlled sizes between 128 and 440 μm and control over the hydrophilic-to-hydrophobic domain volume ratio of the particle from 0.36 to 12.77 for a given size. It is shown that the Janus particle size controls the size of the emulsion drops, thus providing the ability to tune the structure and stability of the resulting emulsions. Stability investigations using centrifugation reveal that particles with the smallest size and a balanced hydrophilic-to-hydrophobic volume ratio (Janus ratio) form emulsions with the greatest stability against coalescence. Particles eventually jam at the interface to form nonspherical droplets. This eﬀect is more pronounced as the hydrogel volume is increased. The large Janus particles permit facile visualization of particle-stabilized emulsions, which result in a better understanding of particle stabilization mechanisms of formed emulsions.

■ INTRODUCTION separation and Oswald ripening. Effective control of the ’ fi Pickering emulsions, or particle-stabilized emulsions, have a emulsifying particle s wetting features is bene cial to achieve wide range of applications in major industries such as, but not enhanced emulsion stability. Janus particles, or particles whose − ff limited to, food, cosmetics, and pharmacy,1 3 where the surface has two distinctly di erent physical properties, are promising candidates for emulsion stabilization. Janus particles traditional surfactant-stabilized emulsions can have adverse ff effects.4 The particles’ enhanced and often irreversible o er the ability to tune the particle contact angle at the oil/ adsorption to the liquid−liquid interface is ideal for creating water interface by changing the hydrophilic/hydrophobic

Downloaded via FLORIDA A & M UNIV on August 29, 2019 at 01:32:2 6 (UTC). domain size ratios, which is not feasible using particles with stable drug-carrying systems and microreactors. For example, fl Pickering emulsion systems have proven to be useful as just a single functionality. Thus, understanding the in uence of the particle architecture on the emulsion stability is necessary templates for microcapsules for drug delivery with biocompat- 8,9 See https://pubs.acs.org/sharingguidelines for options on how t o legitimately share published articles. ible colloidal particles serving as the outer shell of oil-in-water to achieve stable Pickering emulsions, especially for large emulsions.5 They have also been used to form microspheres of droplet sizes. The bicompartmental surface functionalization of amphi- various polymers through controlled polymerization within the 10 enclosed Pickering emulsion boundaries.6 The stability and philic colloidal silica nanoparticles has been reported, where the improved stability (∼15 days) of the particle-stabilized applications of Pickering emulsion drops depend on the ∼ μ emulsifier’s secure attachment to the oil/water interface, and a emulsions ( 30 m) was the result of opposing hydrophilic better understanding of how to maximize the particle’s and hydrophobic sides on the silica particles achieved through adsorption is needed. a tedious synthesis procedure. Although these Janus silica particles were useful, this surface functionalization method Although it is possible to use particles of homogeneous  surface functionality to stabilize emulsions, using particles with allowed no control of the size of the functionalized areas an fi important feature in studying Pickering emulsion stabilization modi ed surfaces of both hydrophilic and hydrophobic parts − affords better stabilization. In some cases, these types of mechanisms and for the control of the particle liquid contact emulsifier particles were formed by simply adding oleic acid to angle. Another attempt at controlling the size and shape of aqueous solutions of hydrophilic silica nanoparticles.7 Although this procedure was simple and reproducible, it Received: January 7, 2019 afforded limited control of the hydrophobicity of the particles, Revised: February 28, 2019 and therefore Pickering emulsions were subject to phase Published: March 6, 2019

Janus particle emulsifiers includes the polymerization of Janus droplets formed from two aqueous solutions of two immiscible monomers.11 However, due to the nature of the synthesis, the synthesized particles were polydispersed with some particles as small as 20 μm and some as large as 100 μm. Flow synthesis in microfluidic devices, however, permits the control of monodispersed Janus droplets as particle templates. Herein, we report stable millimeter-size Pickering emulsions formed using amphiphilic Janus particles. We achieve good control over the overall size of the emulsion droplets by precisely varying the size of the particles. We fabricate the Janus particles in a continuous and facile flow synthesis, which enables the tunability of the particle size from ∼100 to 500 μm with controlled degrees of hydrophobic/hydrophilic portions on each particle. We synthesize the hydrophilic side of the particle as a hydrogel to increase the interaction between the aqueous phase and Janus particle for enhanced adsorption of the emulsifying particle at the oil−water interface. The formed Pickering emulsions are shown to be stable over months at a Figure 1. Fabrication of Janus particles with the hydrophilic domain partial particle surface coverage. The stability of these droplets being a hydrogel and the hydrophobic domain being a crescent-moon is then studied via centrifugation to promote coalescence. shape. (a) Coflow of the TMPTA oil and the PEGDA hydrogel solution through the θ capillary forms Janus drops, where UV exposure induces cross-linkage. (b) Real-time image of the particle ■ EXPERIMENTAL PROCEDURES formation in the glass capillary device. (c) Optical image of the Janus Materials. Poly(ethylene glycol) diacrylate (PEGDA45 wt %, particles with the hydrophilic domain being a hydrogel and the Sigma-Aldrich), 5 wt % ethoxylated trimethylopropane triacrylate hydrophobic domain being a crescent-moon shape. The shape of the γ (ETPTASigma-Aldrich), 1 wt % 2-hydroxy-2-methyl-propiophe- hydrophobic domain depends on the balance of surface tensions, 1, γ γ none (Sigma-Aldrich), and 49 wt % water are mixed together to form 2, and 3. (d) Stabilization of a water drop in the oil phase by the the hydrophilic portion of the inner phase, which we refer to as Janus colloidal surfactant. The hydrogel prefers to submerge in water, “water”. Trimethylopropane triacrylate (TMPTA99 wt %, Sigma- whereas the crescent-moon shape of the particle sits at the curvature Aldrich) and 1 wt % 2-hydroxy-2-methyl-propiophenone are added of the water/oil interface. together to form the hydrophobic portion of the inner phase, which we refer to as “oil”. The oil was colored red using Nile red dye. Silicone oil (95 wt %) and 5 wt % Dow Corning Resin 749 (surfactant) are added together to form the continuous phase. Janus relative centrifugal forces (RCFs) and imaged the size of the emulsion particles are formed using the flow of these solutions in a glass droplets after each run. The procedure continued until coalescence capillary microfluidic device and characterized using confocal occurred. microscopy. Methods. Device fabrication and Janus particle synthesis: a ■ RESULTS AND DISCUSSION circular glass capillary with an outer diameter of 1.5 mm is heated and In this work, we use a simple glass microfluidic device with a pulled using a pipette puller to create a tapered fine orifice of 300 μm θ “θ” two-sided capillary, where tunable interfacial tension for the outlet capillary. A 1.5 mm capillary, with a glass wall between the continuous and inner phases permits the splitting it into two separate channels, is also heated and pulled until fi the opening orifice diameter is 80 μm. The two capillaries are inserted controlled formation of Janus particles. At a xed surfactant concentration, variations of the flow rate ratio of water to oil inside another wider diameter circular glass capillary to form a coaxial ff alignment. The θ capillary ensures that the water and oil come in lead to the formation of amphiphilic Janus particles of di erent contact only at the orifice opening to form Janus droplets. Two hydrophilic-to-hydrophobic domain ratios. We add the tapered 1 mm glass capillaries are inserted into the two compartments particles to a surfactant-free oil and water mixture to form of the θ capillary as inlets for oil and water. Oil and water flow millimeter-size Pickering emulsions, which are stable for up to through the θ capillary to form the inner phase, whereas the silicone a year to date. In addition, through the control of particle size oil flows through the larger outer capillary in the same direction, fl and degree of hydrophobicity, stable Pickering emulsions of resulting in a co ow geometry. Figure 1a,b shows a diagram of the varying droplet sizes and types are formed with the latter device and flow directions. At a continuous phase flow rate of 20 mL/h and an inner-phase demonstrated via catastrophic phase inversion. Due to the ease flow rate (combined oil and water flow rates) of 0.2 mL/h, external of visualization of these particles and the formed emulsion pressure from the high continuous phase flow rate extends the inner drops, this experimental system will give a better under- phase to the jet that breaks into drops due to instability caused by the standing of how surface coverage of particles and orientation at surface tension.12 The droplets are immediately photopolymerized the water/oil interface influence emulsion stability. with a UV light intensity of 25 W/cm2 for 2 s at a 1 cm distance using Formation of Particles: Tuning Particle Size for an Omnicure Series 1000. Emulsification. We show that by increasing the total inner- fi Emulsion Formation. We rst cleaned the Janus particles of the phase flow rate at a constant continuous phase flow rate results silicone oil and surfactant by rinsing in acetone three times. To form in larger particle sizes. The ratio of the oil and hydrogel flow the water-in-oil emulsion drops, 11 mL toluene is first added to the rates in the inner phase is constant at 1:1 to study the effect of acetone-dried particle vial followed by the addition of varying fl amounts of water (0.1, 0.6, or 1 mL) and vortexing for 2 min. Figure only varying the overall ow rates. Thus, the particles discussed 1d shows a schematic of a formed emulsion droplet. in this section exhibit a constant ratio between the hydro- Stability Test. We observed the stability of Pickering emulsions phobic and hydrophilic domains, while the overall particle size using an Allegra X-30R Centrifuge. We spun the samples at different increases.

4694 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article

We keep the continuous phase flow rate constant at 20 mL/ not proportionally increase the size. This suggests that for this h while changing the overall flow rate of the inner phase (oil experimental system one needs to understand the effects of the and hydrogel) to 0.2, 0.4, 0.6, 0.8, and 1 mL/h in a single glass experimental parameters on jet widening and droplet pinch-off device. Using a coflow microfluidic geometry, we form droplets to predict the final particle sizes. where break-off occurs at the end of a jetting stream. Figure 2a Upon the formation of the droplet, the TMPTA oil phase immediately spreads on the surface of the PEGDA hydrogel droplet to reduce the interfacial area with high surface tension between water and silicone oil (continuous phase). This spreading is mitigated by the addition of Dow Corning resin 749 into the silicone oil, where it acts as a surface-acting agent to decrease tension between the continuous phase and water. The relative surface energy of the three fluids determines the amount of spreading and results in an embedded particle for our system, as seen from Figure 1c. The spreading is quantified by the spreading coefficient (S),14 where a negative S value results in particle shapes with one sphere embedded in the other. For S greater than zero, core−shell particles form. Thus, we achieve negative spreading coefficients in our particle system by keeping the resin concentration constant, which gives rise to the observed particle shapes. We therefore achieve complete control of the Janus particle Figure 2. Janus particles of different sizes fabricated by adjusting the size through glass capillary microfluidics. There has been inner-phase flow rate. (a) Confocal images of Janus particles of extensive research on the factors that govern droplet break-off different sizes fabricated at different inner-phase flow rates. The and subsequent particle size as a function of the flow rates. average sizes of the particles from left to right are 128.1, 166.7, 216.8, However, these data are limited to systems where the inner- 270.1, and 311.3 μm. The coefficients of variation (CVstandard fl 13,15−17 ff phase uid is composed of just a single phase. Although deviation/mean size) for the di erent particle sizes are 0.164, 0.226, it follows a trend similar to the single-phase fluid flow, current 0.1908, 0.114, and 0.081 respectively. (b) Linear dependence of the fl work deals with a system in which the inner phase is composed average particle size as a function of the total inner-phase ow rate. fl ff The ratio of the hydrogel to TMPTA oil is constant at 1:1. The error of two uids with di erent viscosities and surface tensions and bars represent standard deviations in particle size. The inset shows the both of these phases are exposed to the continuous phase. The longitudinal length of the particle as the measured average size. physics of such a system (jet widening, droplet pinch-off)is Experimental points are solid squares, whereas the line is a linear fitto not yet clear, and more insight into such systems is therefore the experimental data. The numbers in part (b) correspond to the needed. numbers in part (a). The scale bars are 100 μm. Controlling Particle Domain Sizes. The glass capillary device permits control of the hydrophilic and hydrophobic domains of the Janus particle. We vary the volumes of the shows the resulting cross-linked particles with particle sizes hydrophilic and hydrophobic sides of the Janus particle, also (d ) of 128.1, 166.7, 216.8, 270.1, and 311.3 μm. The p called Janus ratio, by changing the ratio of the flow rates of coefficients of variation (CVstandard deviation/mean size) water to oil while keeping their total flow rate constant. We for the different particle sizes are 0.164, 0.226, 0.1908, 0.114, analyze volumes of the respective domains geometrically by and 0.081, respectively. Figure 2b shows a linear relationship between the flow rate of the inner phase and the Janus droplet treating the Janus particle as a sphere embedded in a spherical size, where the decreasing viscous drag permits larger droplet cap. By dividing the particle into two sections, we obtain three formations and subsequently larger particle sizes. distinct sphere cap volumes that form the hydrophilic and fl hydrophobic portions of the particle. The flow rate of the As expected, we observe that holding the ow rate of the fl continuous phase constant while increasing that of the inner continuous phase is 20 mL/h, whereas the ow rate of the fl inner phase is 0.2 mL/h to achieve a particle size of ∼128 μm. phase serves to generate larger droplets in the micro uidic fl device. While keeping the flow rate of the continuous phase We change the ratio of the ow rates of water to oil to produce constant, an increase in the flow rate of the inner phase into particles with varying Janus ratios. fl the jetting regime leads to large shear at the interface caused by At a 1:1 ratio, the co ow of water and oil forms a jetting ff velocity differences. This results in the deceleration and stream, and droplet break-o occurs in the manner as fl widening of the jetting stream under our experimental described above. At a 7:1 ratio, the co ow of water and oil 13 fi conditions. Undulations in this jetting stream due to forms a jetting stream, and the nal Janus particles have a large Rayleigh−Plateau instability result in a small perturbation in hydrophilic part. Figure 3a shows the resulting Janus particles the cylindrical flow that eventually pinches the jet into produced in a single glass device at flow rate ratios of 0.28, 0.5, droplets. The increase in the droplet size is a result of the 1, 3, and 7. At the smallest flow rate ratio of water to oil, we combination of an increased flow rate and widening of the can see that individual cross-linked particles have a large jetting stream coupled with the pinch-off occurring later. If we volume of oil (red) as the hydrophobic side and a much assume that pinch-off occurs at the same time for different flow smaller volume of water (colorless) as the hydrophilic side. At rates one would expect that the particle size, as measured as the other extreme with a 7:1 ratio, individual cross-linked ∼ 1/3 the length shown in the inset of Figure 2b, scales as dp Qi . particles formed large hydrophilic and small hydrophobic However, analysis of the experimental data in Figure 2b domains. This illustrates the ability of a single glass suggests that increasing the flow rate of the inner phase does microfluidic device to produce particles of varying Janus ratios.

4695 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article

Figure 3. Janus particles with tunable hydrophilic/hydrophobic ratio synthesized in a single device. (a) Confocal images of the Janus particles formed when controlling the hydrophilic-to-hydrophobic domain volume ratios on the cross-linked particle. The red color is the hydrophobic solid domain, whereas the hydrophilic hydrogel domain is colorless. The subsets in each section give a closer look at the Figure 4. Pickering emulsions using different size particle stabilizers. particle. The volume ratios of the hydrophilic domain to hydrophobic (a) Water-in-toluene emulsions stabilized by Janus particles via vortex. domain of the Janus particles from left to right are 0.8, 1.36, 2.67, (b) Average emulsion drop diameter as a function of Janus particle 3.44, and 8.57. The error bars represent standard deviations in the width, dA. The error bars represent standard deviations in emulsion domain volume ratio. All scale bars are 100 μm. (b) Volume ratio of diameter. The plot inset shows the measuring direction of the particle the hydrophilic domain to hydrophobic domain of the Janus particle width used in calculation. (c) Schematic to illustrate the increase in versus the ratio of the inner-phase flow rates (ratio of hydrogel emulsion droplet size when the particle size is increased. solution volume flow rate to monomer oil volume flow rate, where these two fluids form the Janus droplet). Solid squares are the experimental points, whereas the line (linear curve fit to the data) is a or equal to 800 μm in diameter. Because the focus of this work guide to the eye. The numbers correspond to the different particles is on mm-scale droplets, we use 0.8 mm as the threshold. Upon shown in part (a). emulsification of our water−oil−particle system, we observed a dense packing of particles on the sides of the droplets. This is Particles of lower (0.36) and higher (12.77) Janus ratios were also seen in Figure 4a, where the 440 μm particles emulsify the also created using a separate device for further experiments. water but leave free surface on top of the droplet. Although this Figure 3b shows the relationship between the volume is an expected effect for large particles, stable emulsions are still fl domain ratio of Janus particles and the ow rate ratio of water formed. The average surface coverage of our emulsion droplets to oil. Although the relationship between the volume domain across various sized particles can be estimated by relating the fl ratio and ow rate ratio seems close to linear, as seen from the particle size to the droplet diameter. Because the size of graph, we see slight deviations in our experimental results. The Pickering emulsion drops is dependent on the size of the deviation might be due to the evaporation of water, evident in stabilizers, it is possible to predict the droplet size based on our ff the slight hydrogel deformation in Figure 3a. This e ect is synthesized Janus particles. Assuming complete surface cover- more pronounced for the smaller sized Janus particles (128 age at a hexagonal close-packing fraction and a circular particle μ m), the data of which is presented in Figure 3. This is also in cross section at the oil−water interface,20 we can express the accordance with the previous work18,19 in the literature, where emulsion drop diameter (De)as a 1:1 flow rate ratio of the inner phase did not produce a 1:1 24ρ V domain volume ratio. p ML p ]\ De = 0.9 βM ] Pickering Emulsion Formation and Stability. We use ρ M πd 2 ] water A (1) Janus particles synthesized in the previous sections to make N ^ ρ ρ Pickering emulsions in water and oil mixtures. The large size of where p is the particle density (1.1 g/mL), water is the water our particles allows us to observe emulsion drop shapes and density, Vp is the average volume of a single particle, dA sizes without a microscope. These large particles should also (particle width) is the diameter of the widest domain of the afford easier visualization at the water/oil interface when particle that sits above or below the oil/water interface, and β studying the effects of particle orientation on stability. Figure is the mass ratio of emulsified water to Janus particles. A 4a shows millimeter-size water-in-oil (toluene) Pickering similar equation has been used for spherical silica nano- emulsion droplets, using Janus particles of approximately 440 particles, where it was found that a mass ratio of emulsified μm in size. Upon mixing the vial contents (0.55g water, 9.54g fluid to stabilizer as low as 3 yielded smaller emulsions than toluene, 0.19g particles), polydispersed emulsions form, some could be predicted.21 Aiming for smaller emulsions drops, we of which have only partial coverage of the droplet surface by maintain our ratio at 3 for droplet size predictions. When using the Janus particles, as observed by a microscope. The formed eq 1 for 90% surface coverage, we predict drop diameters of emulsions were stable over long periods (approximately a year 2.74, 4.85, and 6.22 mm for particles of 172, 333, and 441 μm, so far) without any coalescence or instability. respectively. These sizes were higher than the diameters We count an emulsion drop as a droplet of the dispersed experimentally observed. Figure 4b shows how the emulsion phase with any number of particles on the surface greater than droplet size changes with the increasing particle size, using a

4696 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article

Figure 5. Stability tests using centrifuge. (a) Average Pickering emulsion drop diameter as a function of relative centrifugal force (RCF), using particles of sizes 172, 333, and 441 μm. The Janus ratio for all particle sizes is about 1.41. (b) Microscopic images of the Pickering emulsions formed by the three different Janus particle sizes before centrifuge (left) and after all RCFs up to 156g (right). The oval shape of the larger emulsions is a result of the centrifuge tube turned on its side to take the image. The error bars represent standard deviations in emulsion diameter. All scale bars are 1 mm. constant β value. At 42% surface coverage, the equation gives smaller sized distribution of Pickering emulsions formed from an almost perfect prediction of droplet size, suggesting that our small particles. Although our form of emulsification is known water droplets are, on average, half covered by the particles to produce polydispersed emulsions,11 smaller particles permit upon initial emulsification. We expect emulsion droplet smaller emulsion droplets with a lower polydispersity. Upon diameter to increase with particle size since large and small initial emulsification, before centrifugation, polydispersity, particles pack at the oil−water interface in the same way, as represented by coefficient of variation (standard deviation/ depicted in Figure 4c. As particle size increases, the particle average), for the 172 μm particles is about 0.02 compared to amount decreases, depleting the total amount of the surface 0.06 for the 441 μm particles. It is not yet clear to us, however, area that can be created. This leads to larger droplets. why the smaller particles would create the more mono- − As others have reported,21 23 our particle-stabilized dispersed system. A systematic study on the particle size, emulsions of partial coverage can remain stable under normal homogenization technique, and interfacial surface area is gravity conditions. We further studied emulsion stability using needed to make a sound conclusion. Figure 5b shows particle a centrifuge to increase the gravitational load on the water size effects on emulsion droplet sizes, before and after droplets. After cleaning and loading cross-linked particles into centrifugation. a centrifuge tube of about 11 mL of toluene, we conducted For Janus particles with a fixed hydrophilic-to-hydrophobic studies on (1) three different sized particles at a fixed Janus ratio, we observe that Pickering emulsions with initial average ratio and (2) three different Janus ratio particles at a fixed size. sizes below 2 mm exhibit the strongest resistance to To compare the emulsion stabilities of the different particle coalescence. Emulsion droplets that are initially larger than types, β in all tests is kept at about 3.0, where the particle’s size this size are likely to destabilize, or they are so large that not is the only changing variable. Once the appropriate amount of much more coalescence can occur. Emulsions formed with 441 water is added, commensurate to the weight of particles, the μm particles form > 2 mm droplets and experience a large tube contents are vortexed for 2 min and imaged to size the degree of coalescence when centrifuged. These initial droplets emulsion droplets. Centrifugation was done on a vortexed are large and partially covered with Janus particles. Initial sample at increasing RCFs (measured relative to the fraction of total particles adsorbed to an interface was roughly acceleration to gravityg force) to observe how the initial estimated (assuming constant droplet size) to be 0.63 for emulsion droplet size increases with RCF. As droplets became droplets formed with 441 μm particles. As RCF increases, past irregularly shaped, the full area was calculated with an image 45g, the size remains relatively constant as most of the water from a top view, whereas droplet depth/height was calculated has already coalesced in one phase. Although these Janus with an image of the side. For the large droplets in Figures 5 particles are effective at stabilizing millimeter-size droplets, and 7, the height imaged from the side was almost the same their large size leaves more uncovered surface before except at the edges where it curves. The diameter of a sphere centrifugation and may be more susceptible to coalescence. with this same volume was then used as the effective diameter Because the rate of size change is faster for larger particle- of the emulsion droplet. Experiments were repeated for a stabilized droplets, observed from Figure 5a, stability must be minimum of three times, and the average diameter with affected by more than just the initial droplet size. In our standard deviation was reported in the figures. experiments performed in the current work, the mass (or Figure 5a shows how the average emulsion droplet diameter volume) of particles and water are kept fixed across the changes with RCF, where the initial droplet sizes, before different sized particles. Hence, the stability of the emulsion for centrifugation, increased with particle size. We also observe a given sized particle can be dependent on other factors, such that the 172 μm particle-stabilized emulsions are almost as surface coverage during centrifugation and the fraction of unaffected by the increase in RCF. We observed a much particles adsorbed and desorbed on and from the droplet

4697 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article

Figure 6. Inversion of water-in-oil Pickering emulsions to oil-in-water emulsions. (a−c) Water-in-oil emulsions with equal amounts of oil and water and stabilized by the Janus particles. The oil phase has a red dye. (d−f) Oil-in-water emulsions after doubling the amount of water in the vial. (g) Water-in-oil droplet covered with particles oriented with the hydrogel into the water and hydrophobic side pointing outward. (h) Phase-inverted oil-in-water droplet covered with particles oriented with the hydrogel pointing outward and the hydrophobic side in the oil. The Janus particles used have an average size of 128 μm.

Figure 7. Stability tests using centrifuge. (a) Average Pickering emulsion drop diameter as a function of relative centrifugal force (RCF), using Janus particles with hydrophilic-to-hydrophobic ratios of 0.51, 1.46, and 12.77. The overall size of the particles is about 323 μm. (b) Magnified version of the graph given in (a) for the particle with a Janus ratio of 1.46 (the diameter increases by almost 50%). (c) Microscopic images of the Pickering emulsions formed by the particles of three different Janus ratios before centrifuge (left) and after all RCFs up to 156g (right). The 12.77 ratio particles stabilizing the emulsion experience interfacial jamming. The error bars represent standard deviations in emulsion diameter. All scale bars are 1 mm. interfaces. Although we conclude that particle size influences added water and vortexing the vial, the type of emulsions flips the initial droplet size, these additional factors will need to be to form oil-in-water emulsions, with the red oil dispersed in a analyzed in tandem to gain a complete understanding of the clear water continuous phase. Figure 6f shows this phase particle and emulsion stability. inversion, where we see red spherical emulsion droplets We also conducted phase inversion experiments on formed suspended in clear water. Some of the oil-in-water drops have emulsions. Figure 6a−c shows pictures of formed water-in-oil air bubbles that pull them to the top of the water. The largest emulsions stabilized by 128 μm particles in equal amounts of droplet in the microscope image for Figure 6c,f is 1.5 mm. the two liquids. As can be seen from the figure, the emulsion Phase inversion is an important qualitative measurement of droplets are clear (water phase), whereas the surrounding oil, Pickering emulsion stability, as most applications require the which has been dyed, is red. After doubling the amount of emulsion drop to retain its type. Because Janus particles can

4698 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article

Figure 8. Bright-field images of different Janus particles sitting at the interface of water in a toluene medium, where the different Janus ratios in the images are (a) 0.38, (b) 1.22, and (c) 10.12. The hydrophobic portion of the Janus particle has a red outline, along with an arrow to show its point of contact with the water/oil interface. have different ratios of hydrophilic-to-hydrophobic domains, 7c shows the largest emulsion drop before and after one can thus tune the wettability of the particle and hence centrifugation to highlight the small change in size. influence phase inversion. Utilizing these Janus particles, we We see that particles with a Janus ratio of 12.77 give the will study how factors such as Janus ratio affect emulsion largest change under external forces, as can be seen in Figure stability. 7a,c. As seen in Figure 7a, emulsions formed with high Janus Janus Ratio Effects on Emulsion Stability. We studied ratio particles immediately formed droplets of nonspherical the emulsion stability using particles with different volume shape before any centrifugation. Upon the application of force, ratios of hydrophilic-to-hydrophobic domains. To understand droplets continue to change shape while increasing in average the results, it is important to consider how the Janus ratio size. Eventually, the droplets form a single structure that affects particle orientation at the oil/water interface. Figure 7 encompasses the full water volume, where continued shows the average emulsion droplet diameter as a function of centrifugation can only change its shape. This could be for a applied RCF along with the bright-field images of the Pickering number of reasons. The shape and hydrophilicity of the emulsionsattwoinstancesintime(1) before any particle may contribute to its instability. Similar to the Bancroft centrifugation and (2) after the application of an RCF of rule for surfactant molecules, emulsion stabilizers that are more 156g. We use particles with Janus ratios of 0.51, 1.46, and hydrophilic tend to form preferred oil-in-water emulsions.26 12.77 in this study. However, when these particles sit at the This means that the particles in the current work with very oil/water interface, the measured particle volume domain ratio large hydrophilic sides may be unstable in the formed water-in- does not fully represent the amount of particle sitting in the oil emulsions. We can visualize this in Figure 8a−c, where we water and the amount in the toluene. Instead, a particle surface see how the Janus ratio affects the interface position of a area ratio is more accurate to indicate the amount of particle particle with hydrophilic-to-hydrophobic volume ratios of 0.38, area covered by water and the area covered by oil. These 1.22, and 10.12, respectively. The hydrophobic portion of the surface area ratios of particle submerged in water to oil are Janus particle has a red outline, along with an arrow to show its 0.27, 1.12, and 13.8 for particles with Janus ratios of 0.51, 1.46, point of contact with the water/oil interface. In studying the and 12.77, respectively. Binks et al.1 reported that the smallest stability of Pickering emulsions using bare silica nanoparticles, and most stable particle-stabilized emulsions occur at the point Ridel et al.20 showed that at constant pH, hydrophilic silica where particles sit at the interface with a contact angle close to particles formed stable oil-in-water Pickering emulsions. As 90 degrees. This is also consistent with the results seen in they increased the volume percent of the dispersed oil phase Figure 7a, where emulsion droplets using particles with a Janus past 60%, they obtained emulsion instability with unemulsified ratio closest to 1.46 show the least response to increases in oil. The hydrophilic silica particles were able to stabilize oil-in- RCF when compared to Janus ratios of the other two extremes. water emulsions when the continuous phase was water, but These 1.46 Janus ratio particle-stabilized droplets still, when the oil phase became continuous, instability ensued, as however, experience a size increase upon centrifugation. Figure the very hydrophilic particles could not form the preferred oil- 7b shows that the droplets stabilized by the 1.46 ratio particles in-water emulsions. almost increase in size by 50% as RCF is increased. The dramatic droplet size changes using particles of 12.77 Nevertheless, this size change is much smaller than those Janus ratio could also be attributed to the lateral compression when using particles of Janus ratios 0.51 and 12.77, and this of the particles’ large hydrogel sides. Interfacial jamming or observation is good evidence that the 90 degree particle tight packing of particles, when stabilizers are packed and contact angle at the oil−water interface indeed enhances the compressed at the interface, occurs immediately before energy required for detachment.24 The almost 1:1 Janus ratio centrifugation for particles of high Janus ratio. This may be a on the particles permits strong absorption to the oil/water result of the cohesive effects of the short dangling polymer interface.25 chains on the hydrogel surfaces, common for hydrogels at From Figure 7a,c, we can see that particles with a Janus ratio liquid−liquid interfaces.27,28 This would essentially create of 0.51 initially form the largest emulsion drops before several irregularly shaped droplets covered with compressed centrifugation compared to those with higher Janus ratios. This particles that encase small volumes of water. Here, hydrogels may be due to the large hydrophobic part of the particle, which inevitably touch surface to surface, as depicted in Figure 8c. prefers the toluene phase to the water phase. This might lead Particles of 0.51 Janus ratio never form nonspherical droplets, to the particles staying in the toluene phase rather than at the whereas those with a Janus ratio of ∼1.41 start to jam at interface, leading to less particles available for emulsification different RCF according to size. These irregularly shaped and hence large emulsion droplets.21 These initially very large droplets, indicative of interfacial jamming, start at RCFs of 156, droplets resist further coalescence. The second row in Figure 94, and 48g for particles of sizes 172, 333, and 441 μm,

4699 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article respectively. It may be that the larger the particle size, the more ■ ACKNOWLEDGMENTS likely for them to concentrate at the bottom of droplet B.H. and S.R. would like to acknowledge funding from NSF surfaces, where it is easier for them to be jammed together. FAMU CREST center award #1735968 for current work. D.W., D.C., and L.-H.C. would like to acknowledge funding ■ CONCLUSIONS from NSF Harvard MRSEC award #DMR 14-20570. We demonstrated the use of an easily fabricated glass capillary microfluidic device to create amphiphilic particles (Janus ■ REFERENCES particles) that can stabilize emulsions in place of surfactants. (1) Binks, B. P.; Fletcher, P. D. I.; Holt, B. L.; Beaussoubre, P.; The device permits full control of the size and structure of the Wong, K. Phase inversion of particle-stabilised perfume oil-water Janus particle by varying the flow rates of the entering fluids. emulsions: experiment and theory. Phys. Chem. Chem. Phys. 2010, 12, We obtained particles of different sizes (128−440 μm) at a 11954−11966. constant flow rate of the continuous phase while increasing the (2) Li, X.; Li, H. Y.; Xiao, Q.; Wang, L. Y.; Wang, M. L.; Lu, X. L.; total flow rate of the inner phase. In addition, when the flow York, P.; Shi, S. L.; Zhang, J. W. Two-way effects of surfactants on rate ratio of water to oil is varied, particles with varying Pickering emulsions stabilized by the self-assembled microcrystals of − hydrophilic and hydrophobic portions were synthesized in a alpha-cyclodextrin and oil. Phys. Chem. Chem. Phys. 2014, 16, 14059 single device (hydrophilic-to-hydrophobic domain volume 14069. ratios of 0.8−8.57). The increase in particle size at a fixed (3) Timgren, A.; Rayner, M.; Sjoo, M.; Dejmek, P. Starch particles for food based Pickering emulsions. Procedia Food Sci. 2011, 1,95− total particle volume diminishes the number of overall particles 103. in the experiment. This subsequently decreases the total (4) Effendy, I.; Maibach, H. I. Surfactants and experimental irritant amount of surface area that can be created, leading to larger contact-dermatitis. Contact Dermatitis 1995, 33, 217−225. droplets. The amphiphilic nature of the particles allows the (5) Wei, Z. J.; Wang, C. Y.; Zou, S. W.; Liu, H.; Tong, Z. Chitosan formation of water-in-oil or oil-in-water emulsions based on nanoparticles as particular emulsifier for preparation of novel pH- whichever fluid acts as the dispersed phase. We demonstrated responsive Pickering emulsions and PLGA microcapsules. Polymer that with increasing size of the Janus particles, the millimeter 2012, 53, 1229−1235. emulsion droplets also increase in size. Emulsion droplets of (6) Zhou, H. O.; Shi, T. J.; Zhou, X. Preparation of polystyrene/ repeatable size and distribution were possible by simple SiO2 microsphere via Pickering emulsion polymerization: Synergistic vortexing, and droplet sizes could be predicted at about half effect of SiO2 concentrations and initiator sorts. Appl. Surf. Sci. 2013, 266,33−38. the maximum particle surface coverage. Induced coalescence (7) Sadeghpour, A.; Pirolt, F.; Glatter, O. Submicrometer-Sized via centrifugation makes particles jam at the interface to form Pickering Emulsions Stabilized by Silica Nanoparticles with Adsorbed nonspherical droplets, where this effect is more pronounced as Oleic Acid. Langmuir 2013, 29, 6004−6012. the hydrogel volume (or Janus ratio) is increased and the (8) Noguchi, T. G.; Iwashita, Y.; Kimura, Y. Dependence of the hydrogel surfaces come into contact. Using Janus particles Internal Structure on Water/Particle Volume Ratio in an Amphiphilic (smallest sizes) with a fixed Janus ratio, emulsions with initial Janus Particle-Water-Oil Ternary System: From Micelle-like Clusters average droplet sizes below 2 mm exhibit the least coalescence to Emulsions of Spherical Droplets. Langmuir 2017, 33, 1030−1036. when centrifuged. The closer particles are to a 1:1 Janus ratio, (9) Kang, C. J.; Honciuc, A. Influence of Geometries on the ff Assembly of Snowman-Shaped Janus Nanoparticles. ACS Nano 2018, the stronger they are against coalescence. Although di erent − from the particle stabilizers commonly used for Pickering 12, 3741 3750. (10) Cao, W.; Huang, R. L.; Qi, W.; Su, R. X.; He, Z. M. Self- emulsions, the use of these larger particles allows the unaided Assembly of Amphiphilic Janus Particles into Monolayer Capsules for visual observation of formed emulsion, which leads to better Enhanced Enzyme Catalysis in Organic Media. ACS Appl. Mater. interpretations on how factors such as the orientations of Interfaces 2015, 7, 465−473. amphiphilic particles at the water/oil interface affect Pickering (11) Wei, D.; Ge, L. L.; Lu, S. H.; Li, J. J.; Guo, R. Janus Particles emulsion stability. Moving forward, our utilization of hydrogels Templated by Janus Emulsions and Application as a Pickering introduces an interesting tool for Pickering emulsion Emulsifier. Langmuir 2017, 33, 5819−5828. applications because of the ease at which we can tune their (12) Umbanhowar, P. B.; Prasad, V.; Weitz, D. A. Monodisperse emulsion generation via drop break off in a coflowing stream. ability to deform at an oil/water interface, swelling nature, and − chemistry, all in a microfluidic device. Further studies on how Langmuir 2000, 16, 347 351. (13) Utada, A. S.; Fernandez-Nieves, A.; Stone, H. A.; Weitz, D. A. factors such as Janus ratio and particle orientation at the Dripping to jetting transitions in coflowing liquid streams. Phys. Rev. water/oil interface using hydrogel amphiphilic particles will be Lett. 2007, 99, No. 094502. useful for understanding Pickering emulsion stability. (14) Kim, S. H.; Abbaspourrad, A.; Weitz, D. A. Amphiphilic Crescent-Moon-Shaped Microparticles Formed by Selective Adsorp- ■ AUTHOR INFORMATION tion of Colloids. J. Am. Chem. Soc. 2011, 133, 5516−5524. (15) Nunes, J. K.; Tsai, S. S. H.; Wan, J.; Stone, H. A. Dripping and Corresponding Author jetting in microfluidic multiphase flows applied to particle and fibre *E-mail: [email protected]. Phone: 001 (850) 410- synthesis. J. Phys. D: Appl. Phys. 2013, 46, No. 114002. 6159. (16) Utada, A. S.; Lorenceau, E.; Link, D. R.; Kaplan, P. D.; Stone, H. A.; Weitz, D. A. Monodisperse double emulsions generated from a ORCID microcapillary device. Science 2005, 308, 537−541. Dong Chen: 0000-0002-8904-9307 (17) Castro-Hernandez,́ E.; Gundabala, V.; Fernandez-Nieves, A.; David Weitz: 0000-0001-6678-5208 Gordillo, J. M. Scaling the drop size in coflow experiments. New J. Subramanian Ramakrishnan: 0000-0001-5108-2198 Phys. 2009, 11, No. 075021. (18) Ge, X. H.; Huang, J. P.; Xu, J. H.; Chen, J.; Luo, G. S. Water-oil Notes Janus emulsions: microfluidic synthesis and morphology design. Soft The authors declare no competing financial interest. Matter 2016, 12, 3425−3430.

4700 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701 Langmuir Article

(19) Xu, K.; Ge, X. H.; Huang, J. P.; Dang, Z. X.; Xu, J. H.; Luo, G. S. A region-selective modified capillary microfluidic device for fabricating water-oil Janus droplets and hydrophilic-hydrophobic anisotropic microparticles. RSC Adv. 2015, 5, 46981−46988. (20) Ridel, L.; Bolzinger, M. A.; Gilon-Delepine, N.; Dugas, P. Y.; Chevalier, Y. Pickering emulsions stabilized by charged nanoparticles. Soft Matter 2016, 12, 7564−7576. (21) Frelichowska, J.; Bolzinger, M. A.; Chevalier, Y. Effects of solid particle content on properties of o/w Pickering emulsions. J. Colloid Interface Sci. 2010, 351, 348−356. (22) Binks, B. P.; Kirkland, M. Interfacial structure of solid-stabilised emulsions studied by scanning electron microscopy. Phys. Chem. Chem. Phys. 2002, 4, 3727−3733. (23) Horozov, T. S.; Binks, B. P. Particle-stabilized emulsions: A bilayer or a bridging monolayer? Angew. Chem., Int. Ed. 2006, 45, 773−776. (24) Chevalier, Y.; Bolzinger, M. A. Emulsions stabilized with solid nanoparticles: Pickering emulsions. Colloids Surf., A 2013, 439,23− 34. (25) Aveyard, R. Can Janus particles give thermodynamically stable Pickering emulsions? Soft Matter 2012, 8, 5233−5240. (26) Binks, B. P. Particles as surfactants - similarities and differences. Curr. Opin. Colloid Interface Sci. 2002, 7,21−41. (27) Li, Z. F.; Ngai, T. Microgel particles at the fluid-fluid interfaces. Nanoscale 2013, 5, 1399−1410. (28) Destribats, M.; Wolfs, M.; Pinaud, F.; Lapeyre, V.; Sellier, E.; Schmitt, V.; Ravaine, V. Pickering Emulsions Stabilized by Soft Microgels: Influence of the Emulsification Process on Particle Interfacial Organization and Emulsion Properties. Langmuir 2013, 29, 12367−12374.

4701 DOI: 10.1021/acs.langmuir.9b00058 Langmuir 2019, 35, 4693−4701